A common legal problem is figuring out how courts synthesize two distinct lines of cases. Using Wexis to research this problem, I did a citing references lookup for Case A and then did a locate within that set for the name of Case B.

This is a clunky solution–but likely the best a lawyer can do with Wexis–because many parties share the same name. So the locate for Case B will inevitably return a whole host of nonsense results which are cases that cite Case A with a plaintiff named Case B who has the exact factual scenario that you don’t care about. Every single time.

The major failure of Wexis in this department is a lack of imagination. Wexis portrays citing references as particular to each case rather than simply as a search.

I can see Wexis mucky mucks envisioning a giant book with all cases that cite Case A. And maybe you can browse through that book for particular words, such as the name of Case B. But they’re still tied to this old-fashioned pulling book off shelves notion of legal research. Perhaps they fail to foresee the dynamic, on-the-fly creation of “books” at the user’s whim. This, of course, is when the user, not a Wexis attorney, begins to index cases according to her needs.

Simplicity of function and the power of search come through abstraction. Citing references should simply be a type of search. The upshot of this observation is that having both a search bar and a citing references link is a redundancy which only serves to clutter the interface. Moreover, where citing references is a type of search, it can easily be combined with other search terms to create more powerful queries. A user then could combine a search for all cases citing Case A and AND that with a search for all cases citing Case B.

Just another painful Wexis problem solved with ease.

My laptop was stolen from the office yesterday while I was at lunch.  What fun would a startup be without any trials and tribulations?

We are lucky to be an AlphaLab startup.  I was still in a state of denial when
Jonathan from gamehuddle had offered to let me borrow a laptop so work could go uninterrupted and today at our weekly meeting the AlphaLab team told me that they will gladly help Sonya Labs cope with the expense of a new laptop. And I thought that I’d have to roll with the punches!

Anecdotes of stolen laptops aside, though, AlphaLab is a truly brilliant program. There is the whole Y-combinator style “incubator” which has been plenty celebrated and applauded in internet circles and all the benefits that come with that. The style is so effective that similar programs have popped up in several cities and AlphaLab offers plenty of that. But, there is one thing that AlphaLab seems to have as a very clear advantage over some of the other similar programs. The other programs seem to be about 20 startups per class working with a team of 6 advisors. Since AlphaLab is a program of Innovation Works. It is actually 6 startups with a team of over 20 experts in all the relevant fields as advisors. Need PR advice? Talk to Terri. Have a quick legal question? Talk to Deborah. Goes without saying that most of the advisors are or have been entrepreneurs themselves.

And there is, of course, the office provided:

Now if only Silas would man his laptop and get some work done instead of taking pictures…

Silas shows us his sensitive side: a languorous shot of his twin muses, Nathan and Rodrigo.

From Copyright, Technology, and Access to the Law (Freedom to Tinker)

The real problem is that statutes often don’t mean jack, because the courts apply 74-pronged tests for this and that to most of those statutes (the vagueness of which seems to be reasonable, and in the defendant’s favor, for most criminal statutes, but they can get extremely difficult to understand and reference for other matters).

For that matter, almost all of common law is nearly completely inaccessible to the layperson surfing the web from home. Much of common law is only ensconced in rulings from the bench that cite precedents going back centuries through our British legal heritage.

While the courts are generally doing a better job than they used to when it comes to making the judiciary accessible to the public - many rulings are posted online, sometimes mere hours after they are handed down - you guys make a great point that it’s the *indexing* that’s really at issue today, more so than the accessibility of today’s rulings. One could almost imagine one day going to Google Law rather than needing to consult a lawyer for every little thing.

(From a comment in this thread.)

Some companies working in the legal research space right now suggest that there’s inherent value in making law open and available to the public. There isn’t much. Law is mostly incoherent to novices, non-lawyers can’t practice law, and most issues that people care about aren’t litigated and decided in a form that will help them. While I agree with most of the comment above, I can’t stay along with his reasoning until the end.

Indexing is crucial, and this relates back to the Problem of Search. However, the main purpose of indexing is not for laypersons but for attorneys. Cheaper access to truth and justice come from the indirect reduction in attorneys’ fees and heightened legal research skills that result from improved legal research tools–not from being able to perform comprehensive legal research on your own as a regular joe. Improved, cheap legal research tools can only result from prying the law loose from the white-knuckled grasp of Lexis and Westlaw. Open law allows for competition. Competition makes everyone (but the duopolists) better off.

The upshot is that Sonya Labs is on a mission to make better, cheaper tools for lawyers–not simply make the law “free”. (Although we appreciate the happy coincidence that lawyers are people, too.) Our orientation, then, is toward the customer, not the ideology.

AlphaLab held an open house yesterday evening for Sonya Labs and the other new portfolio companies. Rodrigo told me that he enjoyed the event, and I’ve yet to hear from Rodrigo or Silas that Nathan embarassed himself or the company. I feel comfortable then declaring the open house a success.

Sonya Labs has been in the news alongside the announcement of AlphaLab:

• Newly formed AlphaLab nurtures technology startups (Pittsburgh Post-Gazette)
• AlphaLab announces class of six hot, promising software startups (Pop City)

… and on, and on, and on.

from Carolina Blawg: Altlaw.org (llauncch.blogspot.com)

AltLaw is a free, open-source legal research tool, an *alt*ernative to Lexis and Westlaw. It provides a database and full-text search for over 700,000 Supreme and Federal Appeals court cases. Take a look!

It’s great to see law librarians take the lead in suggesting alternatives to Lexis and Westlaw. One thing I learned as a summer associate at a major firm is just how helpful law librarians can be. A friend of mine who’s a librarian at Davis Polk & Wardwell said, “Oh yeah, summer associates don’t realize that we can do most of their research for them.”

In many instances, law librarians are gatekeepers to legal research tools and a deep understanding of them. I’m glad that law librarians and legal professionals generally (such as law IT professionals) are adding value by seeking out superior tools for legal research. With the help of these professionals, firms can cut unnecessary costs and attorneys can eliminate the more mundane moments of practicing law.

Via ReadWriteWeb I found the following gem of a quote:

Eleanor Coner, the SPTC’s information officer, said: “Children are very IT-savvy, but they are rubbish at researching. The sad fact is most children these days use libraries for computers, not the books. We accept that as a sign of the times, but schools must teach pupils not to believe everything they read.

“It’s dangerous when the internet is littered with opinion and inaccurate information which could be taken as fact.

This sounds a little bit like someone from the early 1900s being upset that people are learning how to drive and forgetting how to ride horses and citing the fact that roads can be dangerous as evidence.

Sarah makes several good points about this in the RWW article and from one of the links:

One comment on The Scotsman makes a fair, if tired, point:
Easier to blame Wikipedia than the fact that you’re poor parents and your children are out partying or playing video games.
Inaccuracies are found in standard encyclopedias (and newspapers) too. And besides, don’t your schools provide textbooks?

Of course there is good and bad information out there, but that is just as true of printed materials. Critical thinking is a skill, not a property of the communication medium of choice. It should go without saying that as a communication medium the internet is much more powerful, flexible and overall superior than print. Granted, there are still things in print that one can’t find on the intertubes, but that is bound to change in the not-too-distant future when google finishes scanning every book in print out there.

Now, if we could only convince the powers that be that the same applies to case law …

from Web site wants state laws for free (registerguard.com)

Oregon’s laws are undoubtedly public records and available for free at the Legislature’s own Web site, www.leg.state.or.us, Johnson said.

But state law requires that legislative counsel charge to recoup the costs of publishing the statutes — set at $390 for a printed set and $30,000 for the electronic version. The latter sum has been paid by companies such as Lexis-Nexis and Westlaw, which in turn provide searchable databases and charge clients a subscription for their use.

Robin Maxey, a spokesman for Senate President Peter Courtney, said he expects lawmakers to be open-minded about whether it makes sense to uphold the copyright policy in this age of advancing Internet technology.“They want to do the right thing,” he said.

Yet another example of how orthodox strategy is artificially buoyed through law. Of course, orthodox strategy–command & control, closed information, extraction through exclusion–depends on those still mired in orthodox thought, e.g., the Oregon state legislature. And while Oregon’s suit of Public.Resource.org evidences how the old guard exert control, the suit as an obvious last gasp speaks to how orthodox tactics are crumbling, even in law. We at Sonya Labs salute Public.Resource.org, Rodrigo’s most recent post notwithstanding.

It is French for part one. There will be many parts to the series, I’m afraid.

The last couple of weeks have seen us trying to write a parser to get some case data into our own database in our format. The resource.org data looked very clean and structured so “piece of cake” thought Rigo and Nathan. Before we started we had some learning to do: we needed to pick an sql python orm and a library to parse html for us.

I had heard good things about beautiful soup and tried to use it for the same task back in march when we were writing our proof-of-concept prototype. The first time the parser ran it was apparent that it wasn’t fast enough and after some simple timing of the different parts it became obvious that the soup was too slow. So, this time we started by looking at the different python xml libraries and doing some crude timing tests. Following the suggestion of Ian Bicking and because we liked the fact that it had a special-purpose html parser we picked lxml for our job. Fast forward a couple of weeks and it seems that it is fast enough.

As for the orm, we looked a little bit into the usual suspects, but we decided to just use the django orm with a hack that Nathan cooked up to be able to use it with or without the rest of the framework.

Two days went there.

Then we started looking at the data and it seemed straight forward enough: html, all the parts we care about clearly marked with distinctive names and we were off to the races. As soon as we had something that appeared to be working, the first task was to make it paranoic: either it gets the data it expects or dies. Fast forward ten days and our parser was still dying on 90% of the data. It was a big and rich piece of cake, I guess. It turns out that the most regular data set we have is quite irregular and has many exceptions. We still don’t quite understand how this happens, it seems that the aim of resource.org is to store the data in a very uniform way. Somewhere in our TODO is looking at their parsers and seeing if we can identify the problem(s) and give a hand in fixing them.

A full work day after the 90% failure rate we were up to 98% success. Of course, this just means that we made the parser and our format less and less restrictive until we could parse enough data. There is a very clear trade-off there that is fairly general: how much structure you keep vs how much time you spend trying to handle all the nuisances that come with trying to parse that structure.

There are still things to fix, but we decided we could put this a bit closer to the back burner and we moved on to other tasks with the 98% of Supreme Court cases in our database. The F.2d and F.3d volumes of the Federal Reporter come from the same source so soon enough they should also be in our database.